Approximately 30% of patients with clinically localized prostate cancer will develop metastatic disease despite eradication of the local tumor. Presumable, micrometastases were present at the time of initial treatment but were not detected by currently available techniques. It is unlikely, however, that all micrometastatic cells will develop into clinically evident metastatic deposits. Our proposal will determine the importance of micrometastases and the phenotypic changes that occur in these cells that allow for the development of metastatic disease. We will detect micrometastatic cells in the bone marrow and pelvic lymph nodes of patients with prostate cancer by amplifying prostatic specific antigen (PSA) mRNA using the polymerase chain reaction amplification technique. PSA is only present in prostatic epithelial cells. We have successfully identified 1 prostate cancer cell in a population of 10 6 lymphocytes, with no false positive results. We have found prostated cancer micrometastases in 50% of patients with clinically localized prostate cancer. The presence of micrometastases correlates strongly with the pathologic stage of the primary tumor. The development of metastatic cells that produce clinically significant metastases is dependent on a sequence of events that include uncontrolled cell growth, acquisition of the metastatic phenotype, production of growth factors, and active cell growth. To determine the biological potential of micrometastases in patients with prostate cancer, we will ascertain the phenotypic changes that occur in these cells. This will be accomplished by double staining the micrometastases with the PSA antibody (to identify micrometastatic cells) and the following antibodies; p53, PCNA, nm23, retinoblastoma, P120, and TGFbeta-1. We will also evaluate the expression of these genes in the primary tumor using immunohistochemistry and Northern analysis. To determine the prognostic importance of these micrometastases in patients with localized prostate cancer, we will compare the presence, number and phenotype of micrometastases in the lymph nodes and bone marrow with known prognostic factors in prostate cancer, disease-free survival, and site of relapse. We expect to find a constellation of clinical, and scientific parameters that will better stage prostate cancer and identify patients at high risk of disease progression. This will give us the unique ability to identify patients who will not benefit from treatment to the primary tumor. Consequently, we will improve the patient care and quality of life for patients with prostate cancer.